What is going to happen without the mac address? The router will not have the mac address of the machine, that is powered off and probably it will not be on the arp table anymore (default is to hold the address on the table for 6 hours before asking again). So the router needs to know the mac address of the machine to be able to send information.įor the Wake on Lan package, where is the problem? When a package for that machine arrives, it uses the mac address on the table, without asking again. Now he knows the mac address and don´t need to ask for it again(for 6 hours). When the answer arrives, the address goes to the arp table. It asks! So when it receives a package to the ip 192.168.1.45 for the first time, a package is sent via broadcast, and the 192.168.1.45 will answer. What the router do to have this mac? How does it know the mac of every machine? The package arrives, the router converts the ip to mac, and send it to the machine. When a router receives a package ( so routing is already working ok) and needs to send the packet to the machine, it needs to translate the ip to the mac address. I´m going to use a very simple explanation without detailing too much, to be easy to understand the concept. What happens is that people route packages correctly but does not know why they can´t wake on lan the device.Īs arp protocol (address resolution protocol - arp) is done automatically, people don´t have much knowledge on how it works, to proper understand the problem. This means configuring the VPN gateway/finding an option, to forward broadcast traffic from VPN remote clients to the local network.Ģ - Converting the ip to mac address. So routing it is really straightforward, the issue may lie with broadcasting it correctly from the target VPN gateway. As long as the VPN client has the correct routes, it can send a broadcast packet such as 192.168.1.255 (a broadcast address) correctly to the VPN gateway across the internet. Most implementations of the magic packet use UDP port 9 although this really does not matter as long as it is routed correctly and transmitted on the same broadcast domain as the target computer.
So essentially it becomes a matter of getting a regular routable packet to the target host with the "magic" sequence inside its payload. The reason for this is the "magic" sequence can be anywhere within the payload. Yes the WOL magic packet is defined within the constrains of layer 2 but this does not mean it cannot be contained inside a network and transport protocol entity which can then be used to route it across the VPN. Old thread but I wanted to chime in because it is still the top rated search result for "wol over vpn".
0 kommentar(er)
